Google Ads sits closer to intent than almost any other channel — someone typing a search query is often actively looking for a solution right now, which is why it converts well but also why mistakes here cost real money quickly. The foundation is understanding campaign types (Search for intent-driven queries, Display for broader reach, Performance Max for automated cross-network delivery), and how keyword match types control exactly which searches trigger your ad.

Quality Score and ad rank determine both how much you pay and how often you show, and both are influenced by relevance — a tightly themed ad group with matching keywords and ad copy consistently outperforms a loosely organized one, often at a lower cost. Getting the structure right from the start (specific ad groups, relevant keywords, extensions that add real value) matters more than any single bidding trick.

Where Google Ads meets people actively searching, Meta and TikTok ads interrupt people who weren't looking for you — which means the creative itself has to do far more work to earn attention. Targeting still matters (interest-based, lookalike audiences built from existing customers, retargeting people who've already engaged), but a strong audience match with weak creative will still underperform a broad audience with a genuinely scroll-stopping ad.

The platforms also reward different creative instincts: Meta still supports more polished, produced creative across formats like carousels and video, while TikTok rewards content that feels native to the platform — casual, fast-paced, closer to organic content than a traditional ad. Choosing the right format and creative-first approach for the platform's expectation is often the difference between an ad that performs and one that gets scrolled past.

Poor campaign structure is one of the most common and most expensive mistakes in paid advertising — overlapping ad groups competing against each other, keywords or audiences scattered across campaigns with no clear logic, and no clean way to tell which piece of spend is actually driving results. Structure should be built for measurement first: each campaign and ad group organized so you can look at performance data and immediately know what it's telling you.

Bidding strategy should follow from the campaign's actual goal rather than defaulting to whatever the platform suggests. Manual bidding gives more control but demands more attention; automated strategies (CPA or ROAS-based) can perform well once there's enough conversion data to optimize against, but tend to underperform on brand-new campaigns with limited history. Whichever you choose, the discipline that matters most is testing one variable at a time, so you always know what caused a change in performance.

"Automated" bidding is AI in practice: Google's Smart Bidding and fully automated formats like Performance Max use machine learning to adjust bids in real time against signals — device, time of day, location, past conversion patterns — a human couldn't realistically track manually. That's genuinely powerful once the algorithm has enough conversion data to learn from, and genuinely unreliable before that, which is why this course keeps returning to the same rule from Module 1: treat automation as a tool to supervise, not a decision to hand off blind. The same caution applies to AI-generated ad creative and audience expansion tools (Meta's Advantage+ campaigns are the clearest example) — they can produce more variations faster than a human team, but someone still has to judge whether what they produced actually fits the brand and the offer.

The most common budgeting mistake isn't spending too much — it's making a scale-or-kill decision before there's enough data to make it reliably. A test budget exists to answer a specific question (does this audience, offer, and creative combination work at all) before committing meaningfully more money to it, and that question can't be answered honestly with three conversions or two days of data.

Once a campaign has clearly proven itself against defined thresholds — tied to your target CAC and what you know about customer LTV — scaling should be deliberate too: increasing budget gradually (vertical scaling) or expanding to adjacent audiences and creative (horizontal scaling) rather than multiplying spend overnight, which often breaks the very performance you were trying to scale. This lesson is about writing those thresholds down in advance, so scale-or-kill decisions are based on a pre-committed framework rather than in-the-moment emotion.

As a concrete starting point: budget enough to generate at least 30-50 conversions before drawing a scale-or-kill conclusion, or run for a minimum of one full week regardless of spend, whichever takes longer — a campaign that converts on day one can still be a fluke, and one that hasn't converted by day two might just need more time. If you don't know your expected cost-per-conversion yet, a reasonable starting test budget for a first campaign is typically in the $300-$500 range (or your platform's suggested minimum, if higher) — enough to see a real pattern without betting the business on an untested channel.

If you're pre-revenue and don't have real CAC or LTV numbers yet, don't wait for them: set a placeholder target CAC based on what you'd be willing to pay for a customer given your pricing — a common starting rule is one-third of your average order value or first-year contract value — test against that placeholder, and replace it with real data as soon as you have enough sales to calculate an actual number. A rough number you commit to beats no number at all, since it still gives you a threshold to test against instead of judging performance on gut feel.